• [How can the Power of Attribution be developed?
  • Try to verbalize where you look at the problem and think about the solution.
    • Recently, I was looking at my notes during the contest in real time and trying to verbalize them as “thoughts” after the fact.
    • But during the contest, the time pressure is so strong that it’s not really properly verbalized.
    • So look at past questions and try to approach writing a consideration instead of suddenly writing code.
    • Question selection uses “50% probability of correct answer” by AtCoder Problem
    • Compare with official explanations to check for excesses and deficiencies

Try it. - Estimating computational quantities Too many mistakes. - Cases to estimate more:. - ‘I don’t think we’ll be able to do a full search’ - correct answer is full search ABC165C. - Cases where we estimate less:. - “You can do the math with DP” → “DP won’t get it done in time, so we had to attribute it to a much lighter problem by transforming the problem AGC048.

  • Many misunderstandings of problem conditions.
    • They probably realize this by looking at sample data or rubbing it in when they are actually solving the problem. Not writing the code, just the text, has made the mistake more noticeable.
  • It can be done in the bathtub or while washing dishes because a computer is not necessary.
  • It won’t be AC, so it won’t be crossed off the list.
  • When I add a lot of problems (like 30) that I plan to solve together, I get confused because I don’t know which ones I haven’t solved.
  • There are medium difficulty (50% probability of correct answer) and high difficulty (20% probability of correct answer) problems, and the former can be solved quickly (10 min/q), but the fact that it can be solved quickly means that “the thinking parts you already have” are applied, so it is difficult to find new parts from the examples you have collected.
    • And subjectively, I thought, but when I checked how he actually answered the high-difficulty questions, the “questions left unscrambled” weren’t necessarily high-difficulty questions.
      • For example, even the high difficulty problems ABC034C(difficulty 1412) and [ARC106D] are easily solved.
        • I am better at problems involving mathematical transformations than other participants.
      • Conversely, I’m not good at problems that require me to realize that I can solve greedily.
        • When I read ant books and other books, I went through the greedy method with self-knowledge that I “understand” it, but in reality, “I can implement the greedy method after I am sure that it can be solved by the greedy method” and “the part where I realize and prove that it can be solved by the greedy method is weak”, so in the cutoff for the power to be attributed, “I don’t understand”.
  • I’m tired of doing this in a lighthearted way in the difficulty zone.
    • It doesn’t give you a sense of accomplishment.
    • I’ve decided to solve the problems linked from the explanatory article.

done

difficult problem

AGC031B AGC022B AGC032B


This page is auto-translated from /nishio/帰着訓練 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.